Increased Agonistic Behavior in Hummingbirds (Family Trochilidae) in Monteverde, Costa Rica with a Reduction of Food at Artificial Feeders Taegan A. McMahon Department of Biology, Bates College, Lewiston, Maine 04240, USA. ABSTRACT This study looked at the change in agonistic behaviors of hummingbirds (Family Trochilidae) with the reduction of an established food source at la Estacin Biolgica de Monteverde (1567m elevation), Costa Rica. There were three treatment periods: three fee ders (ten days), one feeder (six days; reduction of food source), and three feeders (four days) between October 23 and November 14, 2005. The agonistic behaviors were the number of agonistic behaviors were recorded a well as the species composition of agg ressors versus recipients. Over the three treatment periods there was an increase in hummingbird visits (treatment one to treatment two: p = 0.0030; treatment one to treatment three: p = 0.0005), agonistic behavior when the food source was reduced (p = 0.0 148), and tolerance when the food source was replenished (p = 0.0412). The four dominant aggressors were: Lampornis calolaem, p urple throated mountain gem, Eupherusa eximia, striped tailed hummingbird, Colibri thalassinus green violet ear and Campylopteru s hemileucurus violet sabrewing; the three most targeted recipients were: E. eximia C. thalassinus and C. hemileucurus The artificial feeders increased agonistic behavior such as guarding and darting. This increase potentially occurred because it was more advantageous for the birds to guard the feeders, a constant, rich food source, than to trapline. Such behavior may cause an increase in territoriality and an alteration of the pollination system in areas with established artificial feeder gardens. RESUMEN El propsito de esta investigacin fue determinar el cambio en el comportamiento agresivo de los colibres de la familia Trochilidae con la reduccin de un recurso de alimentario en la Estacin Biolgica de Monteverde (a 1567m de altitud), Costa Rica. Hubo tres perodos de tratamientos: tres comederos (con diez das de duracin), un comedero (con seis das de duracin y con una reduccin de alimento), y tres comederos (con una duracin de cuatro das) del 23 de Octobre al 14 de Noviembre de 2005. Los comportamientos agresivos fueron anotados de acuerdo al nmero de incidentes y a cules pjaros fueron los agresores y cules fueron los recipientes. En los tres tratamientos aumentaron las visitas de los colibres (Del tratamiento uno al dos: p = 0 .0030; del tratamiento uno al tres: p = 0.0005), as como los comportamientos agresivos cuando la fuente de alimiento fue reducida (p = 0.0148) y la tolerancia cuando se reestableci el recurso (p = 0.0412). Los colibres ms dominantes fueron Lampornis ca lolaem, Eupherusa eximia Colibri thalassinus y Campylopterus hemileucurus ; los colibres que recibieron ms comportamientos agresivos fueron E. eximia C. thalassinus y C. hemileucurus. El aumento en los comportamientos agresivos fue causado por los comederos artificiales. Tal vez esto sucedi
porque fue ms beneficioso para los pjaros proteger los comederos, un recurso constante, que trasladarse para alimentarse en otros lugares. Este comportamiento de proteccin pudo causar un aumento en la territorialidad y una alteracin en los sistemas de polinizacin en las reas con comederos artificiales INTRODUCTION Hummingbird (Family Trochilidae) anatomy, behavior, and social interactions are of interest to many, whether they be researching scientists or tourists at an artificial hummingbird garden. All 330 species have similar diet, anatomical features, high wing beat rates, and the ability to hover (Stiles and Skutch 1989). Therefore, n iche partitioning is extremely important. It has allowed for different species to coexist in relatively similar and close habitats, while reducing the need to fight over resources (Stiles and Skutch 1989). Hummingbirds must spend much of their time feedi ng in order to maintain their body temperature, despite efficient (95%) sugar assimilation ( McWhorter and Martinez del Rio 2000) It is thought that hummingbird body size and feeding morphologies have closely coevolved (Altshuler and Dudley 2002; Suarez an d Gass 2002 ). They lose a lot of energy due to their small size and rapid wing beat rate of up to 80 beats/second. Therefore, in order to sustain their high metabolic rate, the highest time averaged metabolic rate among vertebrates ( Voigt and Winter 1999; Suarez and Gass 2002), t hey must change their feeding behavior and social interactions to most efficiently access food ( Altshuler 2004). Stiles and Skutch (1989) state that most hummingbirds are aggressive towards others, regardless of species or sex, at flowers or feeding sites. Social status of the species greatly affects access to nectar sources and quality ( Bleisweiss 1999) ; males are more dominant over conspecific females and species specific hierarchies are often set up at nectar sources (Bleisweiss 1999). These birds often hold feeding territories at consistent feeding locations. There are also trapliners, species that are typically less aggressive and visit many widely spread flowersinstead of maintaining a single permanent feeding territory. Tra plining species are important pollinators because they carry pollen from one flower to another away from the source ( Castellanos et al. 2003) and outcrossing confers fitness for plants. Canela and Sazima (2003) noted that hummingbird feeding behavior coul d change depending on the food source. They found, in a study on the pollination of Aechmea pectinata (Bromeliaceae) in submontane rainforest in southeastern Brazil, that clumped flowering prompted territorial behavior in hummingbirds. Weidner (2001) fou nd that Amazilia tzactl the rufous tailed hummingbird, was extremely aggressive and dominat, thereby reducing the number of other species at feeding sites. Other species, such as Campylopterus hemileucurus the violet sabrewing, are known to be trapliner s and are nonterritorial much of the time (Stiles and Skutch 1989). However, at sites using artificial feeders as the food source C. hemileucurus can be very agonistic (Weidner 2001). This study examined the changes in hummingbird behavior and aggression in response to a reduction in food source. It specifically looked at which species of hummingbirds visited the feeders, their agonistic behaviors, and how their behavior changed with a reduction in food source. It was hypothesized that as the number of f eeders decreased, agonistic behavior between species would increase, and upon food source replenishment tolerance between species would increase. This study showed species species reactions to clumped food sources and identifies species which are more like ly to succeed in food acquisition in areas with reduced food sources;
thereby, increasing the understanding of the effects of artificial feeding gardens on hummingbird behavior. METHODS The data were taken between October 23, 2005 and November 14, 2005 at la Estacin Biolgica de Monteverde, Costa Rica (1567 m elevation) (Fig. 1a). The study site was located adjacent to lower montane wet forest. There were two sites with red feeders (re d attracts hummingbirds) placed in the same small field, which acted as replicates of one another; one was placed three meters from the intact forest edge and the other was placed twenty meters from the same forest edge, next to a large patch of bushes (Fi g. 1b). The feeders were hung two meters off the ground using a blue nylon rope suspended between two trees, and were filled every morning with one liter of 30% sucrose solution each. The feeding sites were set up three days prior to the first treatment. There were three treatment periods: the first test period had three feeders (ten day duration), the second treatment period had one feeder (six day duration), and the third treatment period had three feeders (four days duration). Hummingbird interactions were recorded in one hour increments at 9:00 AM 10:00 AM 11:00 AM and 12:00 PM and the observation period began ten minutes after arrival, to allow hummingbirds to acclimate to the presence of the observer. The number of visits per species was recorded, as were the following agonistic behaviors: Darting, Calling, Striking, Plumage, Ignore and Guarding A Visit was defined as: arrival at the feeder, feeding (any number of times), and leaving the feeder. No agonistic behaviors were considered unless associated with a visit or arrival at a feeder. The agonistic behaviors were qualifi ed as: Darting flying at or chasing another resident or incoming individual, Calling warning calls described by Stiles and Skutch (1989), Plumage ruffling of head and neck feathers in display to another resident or visitor, Ignore a resident individual allowing a visitor to feed, and Guarding perching near the feeder site and confronting visitors. A total of 53 hours of observations were conducted. Treatment period versus total visitations, individual and total agonistic behaviors, were analyzed using
RESULTS The average number of hummingbird visits increased significantly over the three treatment second to thir d treatment p = 0.8606 ; Fig. 2). The total level of agonistic behavior was highest in 0. 6439; second to third treatment: p = 0.0148; Fig. 3), as was the oc currence of guarding events 0524 ; first to third treatment: p = 0. 6796 ; second to third treatment p = 0.0319 ; Fig. 4). There was an overall increase in Ignores over the three t to second treatment: p = 0. 3125 ; first to third treatment: p = 0.0412; second to third treatment: p = 0. 2816 ; Fig. 5). Darting Calling and Plumage were all Darting: f irst to second treatment: p = 0.0176; first to third treatment: p = 0.0125; second to third treatment: p = 0. 0125 ; Calling: first to second treatment: p = 0. 1624 ; first to third treatment: p = 0. 8588 ; second to third treatment: p = 0. 2810; Plumage : first to second treatment: p = 0.0236; first to third treatment: p = 0.0413; second to third treatment: p = 0. 0413 ; Fig. 5). Species that visited the feeders were: C. hemileucurus, Eupherusa eximia, Phaethornis guy, Elvira cupreiceps, Lampornis calolaema, Heliodoxa jacula, Calliphlox bryantae, Selasphorus scintilla, Colibri thalassinus, and Elvira chionura There were four dominant aggressors, written in order of number of agonistic behaviors displayed: L calolaema E. eximia C. thalassinus a nd C. hemileucurus (Fig. 6). E. eximia C. thalassinus and C. hemileucurus were the most targeted species (Fig. 6). E. chionura, the white tailed emerald, a nd C. bryantae the magenta throated woodstar, never acted as aggressors (Fig. 6). ______________________________________________________________________________ Figure 2. The mean number of hummingbird visits to artificial feeders, at la Estacin Biolgica de Monteverde in October and November, 2005, over three treatment periods: treatment one was three feeders (ten days), treatment two was one feeder (six days), and treatment three was three feeders (four days). n = 53. ______________________________________________________________________________ 0 10 20 30 40 50 60 70 One Two Three Number of Visits Treatments
______________________________________________________________________________ Figure 3. The mean number of agonistic events in hummingbirds, at artificial feeders, at la Estacin Biolgica de Monteverde in October and November, 2005, over three treatments treatment one was three feeders (ten days), treatment two was one feeder (six days), and treatment three was three feeders (four days). n = 53. ______________________________________________________________________________ ______________________________________________________________________________ Figure 4. The mean number of guarding events in hummingbirds, at artificial feeders, at la Estacin Biolgica de Monteverde in October and November, 2005, over three treatments: treatment one was three feeders (ten days), treatment two was one feeder (six days), and treatment three was three feeders (four days). n = 53. ______________________________________________________________________________ 0 5 10 15 20 25 30 35 One Two Three Agonistic Actions Treatment 0 1 2 3 4 5 6 7 8 9 One Two Three Guarding Treatments 0 1 2 3 4 5 6 7 8 9 One Two Three Guarding Treatments
_____________ _________________________________________________________________ Figure 5. The mean number of agonistic behaviors (Ignored, Darting, Calling, and Plumage) in hummingbirds at artificial feeders at la Estacin Biolgica de Monteverde in October and November 2005, over three treatment periods: treatment one was ten days (with three feeders), treatment two was six days (with one feeders) and treatment three was four days (with three feeders). n = 53. ______________________________________________________________________________ 0 5 10 15 20 25 Ignored Darting Calling Plumage Number of Actions Behaviors Treatment One Treatment Two Treatment Three
______________________________________________________________________________ Figure 6. The number of aggressor and recipient events between hummingbirds at artificial feeders at la Estacin Biolgica de Monteverde. (a) Number of occurrence, (b) the percentage of agonistic behavior per species. October and November, 2005. n = 53. ___________ ___________________________________________________________________ 0 50 100 150 200 250 Occurances of Agonistic Behavior Species Agonistic Recipient 0 20 40 60 80 100 120 Percentage of Agonistic Behavior Species Percent Agonistic Percent Recipient
DISCUSSION This study showed that hummingbirds increase agonistic behaviors when a constant food source is suddenly limited. The first treatment acted as an acclimatizing period and created a consistent and reliable food source. It was more energetically advantageous for the birds to visit this one ample food source and protect it from others, than it would have been to feed as trapliners, even when the food source was limited. The increase in Guarding in the second treatment represented an increase in territoriality; this is most likely because sharing resources or tolerating other species causes a reduction in fitness when food is limited. The observed increase in tolerance (i ncrease in Ignores ) after the replenishing of the food source further supports the idea that as The average number of visitations increased over all three treatments. This is surprising because a decrease in visitations would be expected in the second treatment with resource reduction. Hypothetically less agonistic species would always be chased feeders in this scenario and would therefore stop wasting energy trying to util ize feeders. The observed increase could be due to gradual learning by hummingbirds that this was a consistent food source. Therefore, the increase in visitation is probably the result of more individuals having learned of the presence of the food source and consequentially more individuals feeding. The increase in territoriality around artificial feeders has important implications for pollination. Castellanos et al. (2003) found that hummingbirds are actually better cross pollinators than other simila r pollinators, such as bees. This is because they are larger and remove more pollen from a given flower. Additionally, removed pollen remains on their body longer, and more pollen is then transferred between flowers. Hence, aggregation leads to a decreas e in pollen transfer, both in distance and occurrence, leading to greater selfing and reduced cross pollination. Aldrich and Hamrick (1998) found an important connection between territoriality and clumped resources. In fragmented forests, where food was m ore clumped, hummingbird territoriality significantly increased. They also found that on trees with larger patches of flowers, such as in pastures, higher territoriality resulted in an increased rate of selfing (Aldrich and Hamrick 1998). Therefore, as f lower patches are reduced, there will be an even larger incidence of territoriality and reduction in traplining. This study showed that an increase in territoriality occurred in some species with the reduction of food. It should not be generalized to all hummingbird species, but territorial hummingbirds will probably respond by narrowing their feeding ranges, which would reduce cross pollination. C. hemileucurus which displays both territorial and traplining behavior, showed high levels of territorialit y at the artificial feeders, therefore reducing their traplining behaviors. The behavioral change may be most drastic in these species, which vary between territoriality and traplining. Suarez and Gass (2002) found that birds showed a preference for high er sucrose food sources, and many the sucrose solutions found in artificial feeders have higher sucrose concentrations than nectar found in flowers. This, coupled with the constancy of artificial feeders, implies that when given the choice between flowers and artificial feeders hummingbirds will choose the artificial feeding gardens. McWhorter and Martinez del Rio (2000) also found that hummingbirds showed a decrease in number of visits when sucrose concentrations were increased. Therefore, high sucrose c oncentrations found in artificial feeders cause both higher territoriality and lower feeding rates, as compared to visiting flowers, therefore reducing frequency of feeding. Presumably, this causes a decrease in pollination in general, specifically cross pollination.
The four most dominant aggressors, L calolaema E. eximia C. thalassinus and C. hemileucurus were assumed to show a higher advantage in situations with fewer feeding sites because they were the dominant species when vying for the limited food source. Interestingly three of these species, E. eximia C. thalassinus and C. hemileucurus were most targeted recipients because they are all extremely territorial and spend much of their time near the feeders. This meant that these species were interspecifically fighting amongst eachother. It should also be noted that when looking at the percentages, r ather than number of interactions, E. chionura and P. guy were most commonly targeted. These two species, according to Stiles and Skutch (1989), are typically trapliners and therefore, they probably only stopped at the feeders when passing between flowers In these instances they were targeted by the more agonistic species. E chionura and C bryantae never acted as aggressors. This was expected for E. chionura because they are very passive, but it was unexpected for C. bryantae Stiles and Skutch (1989 ) described this species as a very agonistic and territorial species. Therefore, the lack of agonistic behavior was probably due to the low number of visits to the feeder. It was also noted that there were very few females visiting the feeders, only two species, E. eximia and L. calolaema, had female visitors. This is probably due to the sexual dimorphism seen between males and females; females are often found at poorer food sources ( Bleisweiss 1999) because males tend to be more dominant and territoria l. This is also evidence that these artificial feeders were considered a good feeding source by the hummingbirds. It was noted that hummingbirds, within one hour of the original feeder set up, had abandoned natural feeding habitats and were instead feedin g solely at artificial feeders. This problem is compounded with the increase in artificial feeding gardens in areas for tourism, which induces territoriality by clumping a constant, high sucrose food source. Therefore, in areas with the artificial feedin g gardens, pollination systems are likely to be disrupted. This is because the more agonistic species, L calolaema E. eximia C. thalassinus and C. hemileucurus which were described by Stiles and Skutch (1989) as either territorial and agonistic or pote ntially territorial, were also the most common visitors. Species described as trapliners (Stiles and Skutch 1989), for example P. guy and E. chionura did not begin to congregate around the feeders as did the agonistic species. Therefore, the biggest imp act on social dynamic and behavior in hummingbirds was on the territorial and more agonistic species, and less, at least to begin with, on the trapliners. These artificial feeders also change the social dynamic within the species because they eliminate t he naturally imposed effects of niche differentiation, because all the birds can feed at the nonspecialized feeders. At these artificial gardens there are interspecific interactions which would not normally have occurred in natural settings, partly becaus e many of these species do not naturally co occur. This changes the system imposed by natural resource availability, which is reduced or eliminated at these artificial feeders. Future research should be done to see how the pollination rates for plants, commonly pollinated by more territorial species, change with distance from artificial feeders. If feeders induce increased territoriality in agonistic species, then they will be e ven less likely to trapline. Three examples of plant species pollinated by several of the above agonistic species were: Heliconia sp. (pollinated by C. hemileucurus), Inga sp. ( pollinated by E. eximia and C. bryantae described as agonistic), and Cephaelis sp. (pollinated by C. hemileucurus and L. calolaema) (Stiles and Skutch 1989). It would also be important to investigate the long term effects on social dynamics at feeder gardens, to see how they are changing the behavior of trapliners and
the less dominant species. Another helpful study would be to study the change in behavioral interactions between species, that do not naturally co occur, at artificial feeding gardens. ACKNOWLEDGEMENTS Thank you to Javier for all of his help and advice To both Ollie and Maria for offering advice on animal avoidance. Thank you to la Estacin Biolgica de Monteverde for housing this experiment, and to everyone who talked and sat with me during my treatment periods. LITERATURE CITED Aldrich, P. and J. Hamrick. 1998. Reproductive dominance of pasture trees in a fragmented tropical forest mosaic. Science. 281:103 105. Altshuler, D., 2004. Of hummingbirds and helicopters: hovering costs, competitive ability, and foraging strategies. Am. Nat. 163(1):16 2 5. Altshuler, D., and R. Dudley. 2002. The ecological and evolutionary interface of hummingbird flight physiology. J Exp. Biol. 205 (16):2325 2336. Bleisweiss. R. 1999. Joint effects of feeding and breeding behaviour on trophic dimorphism in hummingbird s. Proc. Biol. Sci. 226 (1437):4291 4297. Canela, M., and M. Sazima. 2003. Aechmea pectinata : a hummingbird dependent bromeliad with inconspicuous flowers from the rainforest in southeastern Brazil. Ann. Bot. (Lond.). 92 (5):731 737. Castellanos, M., P. Wilson, J. Thomson. 2003. Pollen transfer by hummingbirds and bumblebees, and the divergence of pollination modes in Penstemon. Evolution Int. J. Org. Evolution. 57 (12):2742 2752. McWhorter, T., and C. Martinez del Rio. 2000. Does gut function limit hum mingbird food intake? Physiol. Biochem. Zool. 73 (3):313 324. Stiles, G., and A. Skutch. 1989. A guide to the birds of Costa Rica. pp. 208 231 and Plates 23 25. Comstock Publishing Associates. Ithaca, NY. Suarez, R., C. Gass. 2002. Hummingbird foraging a nd the relation between bioenergetics and behaviour. Comp. Biochem. Physiol. A. Mol. Integr. Physiol. 133 (2):335 343. Voigt, C., Y. Winter. 1999. Energetic cost of hovering flight in nectar feeding bats (Phyllostomidae: Glossophaginae sp.) and its scalin g in moths, birds and bats. J. Comp. Physiol. [B]. 169 (1):38.48. Weidner, K. 2001. Patterns in hummingbird use of a tropical disturbance mosaic. Tropical Biology and Conservation, CIEE. Spring. pp. 309 322.
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xlink http:www.w3.org1999xlink xmlns:xsi http:www.w3.org2001XMLSchema-instance
leader 00000nas 2200000Ka 4500
controlfield tag 008 000000c19749999pautr p s 0 0eng d
datafield ind1 8 ind2 024
subfield code a M39-00209
McMahon, Taegan A.
Aumento en el comportamiento agresivo en los colibres (Familia Trochilidae) en Monteverde, Costa Rica, debido a una reduccin de alimentos en los comederos artificiales
Increased agonistic behavior in hummingbirds (Family Trochilidae) in Monteverde, Costa Rica with a reduction of food at artificial feeders
This study looked at the change in agonistic behaviors of hummingbirds (Family Trochilidae) with the reduction of an established food source at la Estacin Biolgica de Monteverde (1567m elevation), Costa Rica. There were three treatment periods: three feeders (ten days), one feeder (six days; reduction of food source), and three feeders (four days) between October 23 and November 14, 2005. The agonistic behaviors were the number of agonistic behaviors recorded, as well as the species composition of aggressors versus recipients. Over the three treatment periods there was an increase in hummingbird visits (treatment one to treatment two: p = 0.0030; treatment one to treatment three: p = 0.0005), agonistic behavior when the food source was reduced (p = 0.0148), and tolerance when the food source was replenished (p = 0.0412). The four dominant aggressors were: Lampornis calolaem, purple throated mountain gem, Eupherusa eximia, striped tailed hummingbird, Colibri thalassinus, green violet ear and Campylopterus hemileucurus, violet sabrewing; the three most targeted recipients were: E. eximia, C. thalassinus and C. hemileucurus. The artificial feeders increased agonistic behavior such as guarding and darting. This increase potentially occurred because it was more advantageous for the birds to guard the feeders, a constant, rich food source, than to trapline. Such behavior may cause an increase in territoriality and an alteration of the pollination system in areas with established artificial feeder gardens.
El propsito de esta investigacin fue determinar el cambio en el comportamiento agresivo de los colibres de la familia Trochilidae con la reduccin del recurso de alimento en la Estacin Biolgica de Monteverde (a 1567m de altitud), Costa Rica. Hubo tres perodos de tratamientos: tres comederos (con diez das de duracin), un comedero (con seis das de duracin y con una reduccin de alimento), y tres comederos (con una duracin de cuatro das) del 23 de Octubre al 14 de Noviembre del 2005. Los comportamientos agresivos fueron anotados de acuerdo al nmero de incidentes, cules fueron los agresores y cules fueron los recipientes. En los tres tratamientos aumentaron las visitas de los colibres (Del tratamiento uno al dos: p = 0.0030; del tratamiento uno al tres: p = 0.0005), as como los comportamientos agresivos cuando se redujo la fuente de alimento (p = 0.0148) y tolerancia cuando se restableci el recurso (p = 0.0412). Los colibres ms dominantes fueron Lampornis calolaem, Eupherusa eximia, Colibr thalassinus y Campylopterus hemileucurus; los colibres que recibieron ms comportamientos agresivos fueron E. eximia, C. thalassinus, y C. hemileucurus. El aumento en los comportamientos agresivos fue causado por los comederos artificiales. Tal vez esto sucedi porque fue ms beneficioso para las aves proteger a los comederos, un recurso constante, que trasladarse a otros lugares para alimentarse. Este comportamiento de proteccin pudo causar un aumento en la territorialidad y una alteracin en los sistemas de polinizacin en las reas con comederos artificiales.
Text in English.
Agonistic behavior in animals
Monteverde Biological Station (Costa Rica)
Costa Rica--Puntarenas--Monteverde Zone--Monteverde
Conducta agonstica en animales
Estacin Biolgica de Monteverde (Costa Rica)
Costa Rica--Puntarenas--Zona de Monteverde--Monteverde
Tropical Ecology Fall 2005
Ecologa Tropical Otoo 2005
t Monteverde Institute : Tropical Ecology